Project Python Foundations: FoodHub Data Analysis¶

Marks: 60

Context¶

The number of restaurants in New York is increasing day by day. Lots of students and busy professionals rely on those restaurants due to their hectic lifestyles. Online food delivery service is a great option for them. It provides them with good food from their favorite restaurants. A food aggregator company FoodHub offers access to multiple restaurants through a single smartphone app.

The app allows the restaurants to receive a direct online order from a customer. The app assigns a delivery person from the company to pick up the order after it is confirmed by the restaurant. The delivery person then uses the map to reach the restaurant and waits for the food package. Once the food package is handed over to the delivery person, he/she confirms the pick-up in the app and travels to the customer's location to deliver the food. The delivery person confirms the drop-off in the app after delivering the food package to the customer. The customer can rate the order in the app. The food aggregator earns money by collecting a fixed margin of the delivery order from the restaurants.

Objective¶

The food aggregator company has stored the data of the different orders made by the registered customers in their online portal. They want to analyze the data to get a fair idea about the demand of different restaurants which will help them in enhancing their customer experience. Suppose you are hired as a Data Scientist in this company and the Data Science team has shared some of the key questions that need to be answered. Perform the data analysis to find answers to these questions that will help the company to improve the business.

Data Description¶

The data contains the different data related to a food order. The detailed data dictionary is given below.

Data Dictionary¶

  • order_id: Unique ID of the order
  • customer_id: ID of the customer who ordered the food
  • restaurant_name: Name of the restaurant
  • cuisine_type: Cuisine ordered by the customer
  • cost: Cost of the order
  • day_of_the_week: Indicates whether the order is placed on a weekday or weekend (The weekday is from Monday to Friday and the weekend is Saturday and Sunday)
  • rating: Rating given by the customer out of 5
  • food_preparation_time: Time (in minutes) taken by the restaurant to prepare the food. This is calculated by taking the difference between the timestamps of the restaurant's order confirmation and the delivery person's pick-up confirmation.
  • delivery_time: Time (in minutes) taken by the delivery person to deliver the food package. This is calculated by taking the difference between the timestamps of the delivery person's pick-up confirmation and drop-off information

Let us start by importing the required libraries¶

In [1]:
# import libraries for data manipulation
import numpy as np
import pandas as pd

# import libraries for data visualization
import matplotlib.pyplot as plt
import seaborn as sns
!pip install plotly
import plotly.express as px

import warnings
warnings.filterwarnings("ignore")
pd.set_option("display.float_format", lambda x: "%.2f" % x) #to display values upto 2 decimal places

Requirement already satisfied: plotly in c:\users\tayo adeyo\anaconda3\lib\site-packages (5.6.0)
Requirement already satisfied: six in c:\users\tayo adeyo\anaconda3\lib\site-packages (from plotly) (1.16.0)
Requirement already satisfied: tenacity>=6.2.0 in c:\users\tayo adeyo\anaconda3\lib\site-packages (from plotly) (8.0.1)

Understanding the structure of the data¶

In [2]:
# read the data
df = pd.read_csv('C:\\Users\\Tayo Adeyo\\Downloads\\foodhub_order.csv')
# returns the first 5 rows
df.head()
Out[2]:
order_id customer_id restaurant_name cuisine_type cost_of_the_order day_of_the_week rating food_preparation_time delivery_time
0 1477147 337525 Hangawi Korean 30.75 Weekend Not given 25 20
1 1477685 358141 Blue Ribbon Sushi Izakaya Japanese 12.08 Weekend Not given 25 23
2 1477070 66393 Cafe Habana Mexican 12.23 Weekday 5 23 28
3 1477334 106968 Blue Ribbon Fried Chicken American 29.20 Weekend 3 25 15
4 1478249 76942 Dirty Bird to Go American 11.59 Weekday 4 25 24

Observations:¶

The DataFrame has 9 columns as mentioned in the Data Dictionary. Data in each row corresponds to the order placed by a customer.

Question 1: How many rows and columns are present in the data? [0.5 mark]¶

In [3]:
# No of rows and columns contained in the dataset
df.shape
Out[3]:
(1898, 9)

Observations:¶

There are 1898 rows and 9 columns in the dataset

Question 2: What are the datatypes of the different columns in the dataset? (The info() function can be used) [0.5 mark]¶

In [4]:
# Use info() to print a concise summary of the DataFrame
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1898 entries, 0 to 1897
Data columns (total 9 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   order_id               1898 non-null   int64  
 1   customer_id            1898 non-null   int64  
 2   restaurant_name        1898 non-null   object 
 3   cuisine_type           1898 non-null   object 
 4   cost_of_the_order      1898 non-null   float64
 5   day_of_the_week        1898 non-null   object 
 6   rating                 1898 non-null   object 
 7   food_preparation_time  1898 non-null   int64  
 8   delivery_time          1898 non-null   int64  
dtypes: float64(1), int64(4), object(4)
memory usage: 133.6+ KB

Observations:¶

  • All the colums have the same number of observations - 1898, which tend to suggest that we dont have any missing entry.

  • The "restaurant_name", "cuisine_type", and the "day_of_the_week" columns are categorical variables and are being correctly read as object types.

  • The "food_preparation_time" and "delivery_time" are numerical variables and are being read as integers

  • The "cost_of_the_order" is also a numerical variable and is being read as a float

  • The "order_id" and the "customer_id" are being read in as integers, but these are categorical variables and should be object types, but we'll leave the type as it is.

  • The rating is being wrongly read in as an object type. This is a numerical variable that should be read in as an integer. This suggests that there must be something wrong with some of the entries in this column. There is the need for further investigation of this column.

Question 3: Are there any missing values in the data? If yes, treat them using an appropriate method. [1 mark]¶

In [5]:
# Find the different unique entries in the "rating" column
df.rating.unique()
Out[5]:
array(['Not given', '5', '3', '4'], dtype=object)
In [6]:
# How many observations in the "rating" column has the entry as "Not given"
(df.rating=="Not given").value_counts()
Out[6]:
False    1162
True      736
Name: rating, dtype: int64
In [7]:
# Replace the "Not given" entries in the "rating" column with NaN
df['rating'] = df['rating'].replace(['Not given'],np.nan)
df.head()
Out[7]:
order_id customer_id restaurant_name cuisine_type cost_of_the_order day_of_the_week rating food_preparation_time delivery_time
0 1477147 337525 Hangawi Korean 30.75 Weekend NaN 25 20
1 1477685 358141 Blue Ribbon Sushi Izakaya Japanese 12.08 Weekend NaN 25 23
2 1477070 66393 Cafe Habana Mexican 12.23 Weekday 5 23 28
3 1477334 106968 Blue Ribbon Fried Chicken American 29.20 Weekend 3 25 15
4 1478249 76942 Dirty Bird to Go American 11.59 Weekday 4 25 24
In [8]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1898 entries, 0 to 1897
Data columns (total 9 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   order_id               1898 non-null   int64  
 1   customer_id            1898 non-null   int64  
 2   restaurant_name        1898 non-null   object 
 3   cuisine_type           1898 non-null   object 
 4   cost_of_the_order      1898 non-null   float64
 5   day_of_the_week        1898 non-null   object 
 6   rating                 1162 non-null   object 
 7   food_preparation_time  1898 non-null   int64  
 8   delivery_time          1898 non-null   int64  
dtypes: float64(1), int64(4), object(4)
memory usage: 133.6+ KB
In [9]:
# changing the data type of the "rating" column to type Int64
df['rating'] = df['rating'].astype('Int64')
df.head()
Out[9]:
order_id customer_id restaurant_name cuisine_type cost_of_the_order day_of_the_week rating food_preparation_time delivery_time
0 1477147 337525 Hangawi Korean 30.75 Weekend <NA> 25 20
1 1477685 358141 Blue Ribbon Sushi Izakaya Japanese 12.08 Weekend <NA> 25 23
2 1477070 66393 Cafe Habana Mexican 12.23 Weekday 5 23 28
3 1477334 106968 Blue Ribbon Fried Chicken American 29.20 Weekend 3 25 15
4 1478249 76942 Dirty Bird to Go American 11.59 Weekday 4 25 24
In [10]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1898 entries, 0 to 1897
Data columns (total 9 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   order_id               1898 non-null   int64  
 1   customer_id            1898 non-null   int64  
 2   restaurant_name        1898 non-null   object 
 3   cuisine_type           1898 non-null   object 
 4   cost_of_the_order      1898 non-null   float64
 5   day_of_the_week        1898 non-null   object 
 6   rating                 1162 non-null   Int64  
 7   food_preparation_time  1898 non-null   int64  
 8   delivery_time          1898 non-null   int64  
dtypes: Int64(1), float64(1), int64(4), object(3)
memory usage: 135.4+ KB

Observations:¶

  • There are no missing entries in the dataset provided as each column contains 1898 non-null values

  • There is nevertheless something amiss with the "rating" column since the data type is of the wrong type.

  • On close observation, we find many, (736 in all) were not given any rating, hence the data type - object

  • These entries were changed to "NaN", and the data type converted to 'Int64'

  • The "rating" column now reflects the fact that only 1162 of the entries have non null values, while 736 of the entries now have null values

Question 4: Check the statistical summary of the data. What is the minimum, average, and maximum time it takes for food to be prepared once an order is placed? [2 marks]¶

In [11]:
# Statistical summary of the dataset
df.describe(include='all').T
Out[11]:
count unique top freq mean std min 25% 50% 75% max
order_id 1898.00 NaN NaN NaN 1477495.50 548.05 1476547.00 1477021.25 1477495.50 1477969.75 1478444.00
customer_id 1898.00 NaN NaN NaN 171168.48 113698.14 1311.00 77787.75 128600.00 270525.00 405334.00
restaurant_name 1898 178 Shake Shack 219 NaN NaN NaN NaN NaN NaN NaN
cuisine_type 1898 14 American 584 NaN NaN NaN NaN NaN NaN NaN
cost_of_the_order 1898.00 NaN NaN NaN 16.50 7.48 4.47 12.08 14.14 22.30 35.41
day_of_the_week 1898 2 Weekend 1351 NaN NaN NaN NaN NaN NaN NaN
rating 1162.00 NaN NaN NaN 4.34 0.74 3.00 4.00 5.00 5.00 5.00
food_preparation_time 1898.00 NaN NaN NaN 27.37 4.63 20.00 23.00 27.00 31.00 35.00
delivery_time 1898.00 NaN NaN NaN 24.16 4.97 15.00 20.00 25.00 28.00 33.00

Observations:¶

  • It takes a minimum of 20 minutes and a maximum of 35 minutes for the food to be prepared after the order is placed.

  • It takes about 27.37 minutes on the average to prepare the food after placing an order. It also takes about 24.16 minutes on the average to deliver the food after preparation.

  • It takes a minimum of 15 minutes and a maximum of 33 minutes for the food to be delivered after preparation.

  • Shake Shack is the most patronised out of the 178 restaurant in the dataset with a total order of 219 out of the 1898 orders placed and delivered.

  • there are more orders on weekends than weekdays, 1351 out of the total 1898 orders.

  • American." cuisine is the most favoured among the 14 different cuisine types being responible for a whopping 584 of the orders for food placed.

  • Whenever a rating is given, the minimum rating received by any order is 3 and the maximum is 5 with more than a third (736 out of 1898) of the deliveries made not receiving any rating at all.

Question 5: How many orders are not rated? [1 mark]¶

In [12]:
# Display the number of null entries in the "rating" column (ratings with "Not given" entries has been changed to Null in Q3)
df.rating.isnull().sum()
Out[12]:
736

Observations:¶

736 out of 1898 orders were not rated. Ratings with "Not given" entries has been replaced with Null entries.

Exploratory Data Analysis (EDA)¶

Univariate Analysis¶

Question 6: Explore all the variables and provide observations on their distributions. (Generally, histograms, boxplots, countplots, etc. are used for univariate exploration.) [9 marks]¶

In [13]:
sns.histplot(data=df, x='cost_of_the_order', kde=True);
plt.show()
sns.boxplot(data=df, x='cost_of_the_order');
plt.show()

Observations:¶

  • There are no outliers in the cost_of_the_order

  • 50% of the orders placed cost less than $15

  • The distribution is right skewed

In [14]:
sns.histplot(data=df, x='rating');
plt.show()
sns.boxplot(data=df, x='rating');
plt.show()

Observations:¶

  • 50% (median) of the orders placed and delivered (when rated) were giving the maximum rating of 5

  • The distribution is left skewed

In [15]:
sns.histplot(data=df, x='food_preparation_time');
plt.show()
sns.boxplot(data=df, x='food_preparation_time');
plt.show()

Observations:¶

  • There are no outliers

  • It takes a minimum of 20 minutes and a maximum of 35 minutes to prepare any type of cuisine ordered with roughly about 27 minutes being the average time for food preparation.

  • The distribution is normal

In [16]:
sns.histplot(data=df, x='delivery_time');
plt.show()
sns.boxplot(data=df, x='delivery_time');
plt.show()

Observations:¶

  • There are no outliers
  • 50% of the orders for food are delivered 25 minutes after preparation
  • The distribution is slightly left skewed
In [17]:
# Find the number of unique occurences of order_id
df.order_id.nunique()
Out[17]:
1898

Observations:¶

Even though order_id is of type integer, it is actually a categorical variable and not a numerical variable. there are 1898 unique order_id attached to each order placed and delivered.

In [18]:
# Find the number of customers that has patronised FoodHub
df.customer_id.nunique()
Out[18]:
1200

Observations:¶

Customer_id is also a categorical variable reprenting every unique customer that has utilized the services of FoodHub at one time or the other. There are 1200 unique customers in the dataset

In [19]:
# Find the number of restaurant utilizing the services of FoodHub
df.restaurant_name.nunique()
Out[19]:
178
In [20]:
his = px.histogram(df, x="restaurant_name")
his.show()

Observations:¶

There are 178 different restaurants being serviced by FoodHub with Shake Shack enjoying the most patronage from the registered customers followed by The Meatball Shop and Blue Ribbon Sushi.

In [21]:
# Find the different cuisine types catered for by the restaaurants
df.cuisine_type.nunique()
Out[21]:
14
In [22]:
his = px.histogram(df, x="cuisine_type")
his.show()

Observations:¶

There are 14 different cuisine types being catered by the different restaurants. From the graphical display, the American cuisine seems to be the most requested for cuisine type followed by japanese and Chinese cuisines. Bringing up the rear in the order of prefereed cuisines is Vietnamese cuisine, closely followed by Spanish and Korean cuisines.

In [23]:
# Find unique days of the week
df['day_of_the_week'].nunique()
Out[23]:
2
In [24]:
sns.countplot(data=df, x='day_of_the_week');

Observations:¶

There are two unique days of the week. Weekdays and weekends

Question 7: Which are the top 5 restaurants in terms of the number of orders received? [1 mark]¶

In [25]:
df['restaurant_name'].value_counts().iloc[:5]
Out[25]:
Shake Shack                  219
The Meatball Shop            132
Blue Ribbon Sushi            119
Blue Ribbon Fried Chicken     96
Parm                          68
Name: restaurant_name, dtype: int64

Observations:¶

The Shake Shack restaurant services the highest number of orders at 219 orders. The Meatball Shop and Blue Ribbon Shushi comes a distant second and third with 132 and 119 orders respectively. Blue Ribbon Fried Chicken and Parm is no 4 and 5 with 96 and 68 orders respectively.

Question 8: Which is the most popular cuisine on weekends? [1 mark]¶

In [26]:
# To find the most popular cuisine on weekends
df[df['day_of_the_week']=='Weekend'].cuisine_type.value_counts()
Out[26]:
American          415
Japanese          335
Italian           207
Chinese           163
Mexican            53
Indian             49
Mediterranean      32
Middle Eastern     32
Thai               15
French             13
Korean             11
Southern           11
Spanish            11
Vietnamese          4
Name: cuisine_type, dtype: int64
In [27]:
df[df['day_of_the_week']=='Weekend'].cuisine_type.value_counts().iloc[:1]
Out[27]:
American    415
Name: cuisine_type, dtype: int64
In [28]:
# This can also be done graphically
plt.figure(figsize=(15,7))
sns.countplot(data=df, x='cuisine_type',hue='day_of_the_week');
plt.xticks(rotation = 45);

Observations:¶

The most popular cuisine ordered on weekends is American with 415 orders placed, followed by Japanese with 335 with Italian and Chinese menu taking the 3rd and 4th place with 207 and 163 orders respectively. Incidentally, the prefered cuisines on weekdays also follows the same order

Question 9: What percentage of the orders cost more than 20 dollars? [2 marks]¶

In [29]:
count=0
# find the number of rows where cost_of_the_order is greater than 20 dollars
for i in range(df.shape[0]):   # df.shape[0] is the number of rows in the dataframe
    if df.cost_of_the_order[i]>20:
        count=count+1
    else:
        count=count 
pct_20=(count/df.shape[0])*100
print(pct_20,'%')
print(count, 'of the orders placed cost more than 20 dollars')
print('This is about',round(pct_20, 2),'percent of the total orders placed')

29.24130663856691 %
555 of the orders placed cost more than 20 dollars
This is about 29.24 percent of the total orders placed

Observations:¶

Out of a total of 1898 orders placed and delivered, 555 of these orders cost more than 20 dollars. This is about 29.24 percent of the total orders placed and delivered.

Question 10: What is the mean order delivery time? [1 mark]¶

In [30]:
# Find the total time taken to deliver the order (from when the order was placed, to when it is delivered
df['total_delivery_time'] = df['food_preparation_time'] + df['delivery_time']
mdt=df.total_delivery_time.mean()
print('The mean delivery time for orders is',round(mdt,2),'minutes')

The mean delivery time for orders is 51.53 minutes

Observations:¶

By order delivery time, I take this to mean the time that will elapse between the customer placing an order and the food being delivered to the customer. The delivery time for orders placed is 51.53 minutes on the average

Question 11: The company has decided to give 20% discount vouchers to the top 3 most frequent customers. Find the IDs of these customers and the number of orders they placed. [1 mark]¶

In [31]:
# Customer ids of the top 3 customers and amount of orders placed
df.customer_id.value_counts().iloc[:3]
Out[31]:
52832    13
47440    10
83287     9
Name: customer_id, dtype: int64

Observations:¶

The top 3 customers with the highest number of orders placed are customers with: customer_id: 52832 with 13 orders customer_id: 47440 with 10 orders and customer_id: 83287 with 9 orders

Multivariate Analysis¶

Question 12: Perform a multivariate analysis to explore relationships between the important variables in the dataset. (It is a good idea to explore relations between numerical variables as well as relations between numerical and categorical variables) [10 marks]¶

In [32]:
num_var=['cost_of_the_order','rating','food_preparation_time','delivery_time']
sns.heatmap(df[num_var].corr(), annot=True, vmin=-1, vmax=1, fmt=".2f", cmap="Spectral");
In [33]:
sns.pairplot(data=df[num_var], diag_kind="kde");

Observation:¶

I really cannot see any correllation whether positive or negative between any of the numeric variables from the heatmap and the pairplot displayed above.

In [34]:
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='rating')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='rating',hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()

Observation:¶

The rating accross all cuisine types is similar, whether the rating is being done on weekdays, or on weekends except for a few abberations like Korean, mediterranian, southern and french cuisines that receives dissimilar rationgs on weekdays and weekends.

In [35]:
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='food_preparation_time',hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,6))
sns.pointplot(data=df, x='cuisine_type', y='food_preparation_time', hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
In [36]:
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='delivery_time',hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,6))
sns.pointplot(data=df, x='cuisine_type', y='delivery_time', hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
In [37]:
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='total_delivery_time',hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,6))
sns.pointplot(data=df, x='cuisine_type', y='total_delivery_time', hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
In [38]:
# Number of orders placed on weekdays and weekends
df.day_of_the_week.value_counts()
Out[38]:
Weekend    1351
Weekday     547
Name: day_of_the_week, dtype: int64

Observation¶

  • preparing each cuisine type takes almost the same amount of time whether done on weekdays or weekends. One would have expect preparation time to be longer on weekends due to the volume of orders (over 70% of orders are placed on weekeend), but this is not so.
  • It takes longer on the average to deliver orders placed on weekdays than those placed on weekends. This is to be expected because there is always the likelihood of traffic being lighter on weekends
In [39]:
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='cost_of_the_order')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,4)) 
sns.boxplot(data=df, x='cuisine_type', y='cost_of_the_order',hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,6))
sns.pointplot(data=df, x='cuisine_type', y='cost_of_the_order', hue='day_of_the_week')
plt.xticks(rotation=45);
plt.show()

Observation:¶

  • Most customers tend to buy less expensive meals irrespective of the cuisine type

  • The exception to the above are French and Thai cuisines where more of those who patronised them tend to order the more expensive meals.
    These two insights could be deduced from the skewness of the plots for the differentb cuisine types.

  • The average cost of food differs on weekends from what its average on weekdays. The exception to this are Japanese,Indian and southern cuisine types where the average prices are the same or almost the same on weekdays and weekends. Italian, Chinese, middle eastern, and French cuisines are on the average more expensive on weekdays than on weekends, wheras the reverse is the case for Korean, Mexican, american, Mediterranian, Thai, Spanish amnd Vietnamese cuisines

In [40]:
df2=df.cuisine_type.value_counts().reset_index()
df2 = df2.rename(columns={'index': 'cuisine_type','cuisine_type':'no_of_orders'})
df2
Out[40]:
cuisine_type no_of_orders
0 American 584
1 Japanese 470
2 Italian 298
3 Chinese 215
4 Mexican 77
5 Indian 73
6 Middle Eastern 49
7 Mediterranean 46
8 Thai 19
9 French 18
10 Southern 17
11 Korean 13
12 Spanish 12
13 Vietnamese 7
In [41]:
df3 = df.groupby(['cuisine_type'])['cost_of_the_order'].mean().reset_index()
df3 = df3.rename(columns={'cost_of_the_order': 'ave_cost_of_each_order'})
df3
Out[41]:
cuisine_type ave_cost_of_each_order
0 American 16.32
1 Chinese 16.31
2 French 19.79
3 Indian 16.92
4 Italian 16.42
5 Japanese 16.30
6 Korean 14.00
7 Mediterranean 15.47
8 Mexican 16.93
9 Middle Eastern 18.82
10 Southern 19.30
11 Spanish 18.99
12 Thai 19.21
13 Vietnamese 12.88
In [42]:
df4 = df.groupby(['cuisine_type'])['cost_of_the_order'].sum().reset_index()
df4 = df4.rename(columns={'cost_of_the_order': 'total_cost_of_all_orders'})
df4
Out[42]:
cuisine_type total_cost_of_all_orders
0 American 9530.78
1 Chinese 3505.62
2 French 356.29
3 Indian 1235.14
4 Italian 4892.77
5 Japanese 7663.13
6 Korean 182.02
7 Mediterranean 711.84
8 Mexican 1303.85
9 Middle Eastern 922.21
10 Southern 328.11
11 Spanish 227.93
12 Thai 364.95
13 Vietnamese 90.18
In [43]:
df5 = df.groupby(['cuisine_type'])['rating'].mean().reset_index()
df5 = df5.rename(columns={'rating': 'ave_rating'})
df5
Out[43]:
cuisine_type ave_rating
0 American 4.30
1 Chinese 4.34
2 French 4.30
3 Indian 4.54
4 Italian 4.36
5 Japanese 4.37
6 Korean 4.11
7 Mediterranean 4.22
8 Mexican 4.42
9 Middle Eastern 4.24
10 Southern 4.31
11 Spanish 4.83
12 Thai 4.67
13 Vietnamese 4.00
In [44]:
df6=pd.merge(df2, df3, on='cuisine_type')
df7=pd.merge(df4, df5, on='cuisine_type')

df_cuisine=pd.merge(df6, df7, on='cuisine_type')
df_cuisine
Out[44]:
cuisine_type no_of_orders ave_cost_of_each_order total_cost_of_all_orders ave_rating
0 American 584 16.32 9530.78 4.30
1 Japanese 470 16.30 7663.13 4.37
2 Italian 298 16.42 4892.77 4.36
3 Chinese 215 16.31 3505.62 4.34
4 Mexican 77 16.93 1303.85 4.42
5 Indian 73 16.92 1235.14 4.54
6 Middle Eastern 49 18.82 922.21 4.24
7 Mediterranean 46 15.47 711.84 4.22
8 Thai 19 19.21 364.95 4.67
9 French 18 19.79 356.29 4.30
10 Southern 17 19.30 328.11 4.31
11 Korean 13 14.00 182.02 4.11
12 Spanish 12 18.99 227.93 4.83
13 Vietnamese 7 12.88 90.18 4.00
In [45]:
plt.figure(figsize=(15,5)) 
sns.barplot(data=df_cuisine, x='cuisine_type', y='ave_rating')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,5)) 
sns.barplot(data=df_cuisine, x='cuisine_type', y='no_of_orders')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,5)) 
sns.barplot(data=df_cuisine, x='cuisine_type', y='ave_cost_of_each_order')
plt.xticks(rotation=45);
plt.show()
plt.figure(figsize=(15,5)) 
sns.barplot(data=df_cuisine, x='cuisine_type', y='total_cost_of_all_orders')
plt.xticks(rotation=45);
plt.show()
In [46]:
# Total amount in dollars generated from selling all cuisine types
round(df_cuisine.total_cost_of_all_orders.sum(),2)
Out[46]:
31314.82
In [47]:
# Amount in dollars made from selling the 4 topmost cuisine types
round(df_cuisine.total_cost_of_all_orders.head(4).sum(),2)
Out[47]:
25592.3

Observation:¶

  • American cuisines generate the highest income for all the restaurants at almost 10,000 dollars. This is followed by Japanese cuisines at almost 7,500 dollars and italian and Chinese cuisines at almost 5,000 and 3,500 dollars respectively. bringing up the rear are Vietnamese, Korean and Spanish cuisines at 90, 180 and 230 dollars respectively. This trend is comparable to the number of orders received for each cuisine type.

  • 4 out of the total 14 cuisine type dealt with (28.57%) are responsible for 1597 out of the total 1898 (82.56%) orders placed (20/80 rule?).

  • On the average, Middle Eastern, Thai, French, Southern and Spanish cuisines cost more than the other type of cuisines.

  • Spanish cuisine has the highest average rating of all the cuisine types followed by Thai and indian cuisines respectively.

Question 13: The company wants to provide a promotional offer in the advertisement of the restaurants. The condition to get the offer is that the restaurants must have a rating count of more than 50 and the average rating should be greater than 4. Find the restaurants fulfilling the criteria to get the promotional offer. [3 marks]¶

In [48]:
# Filter out the rated restaurants as those with nonnull rating entries 
df_rated_restaurants = df[df.rating.notnull()].copy()
df_rated_restaurants['rating'] = df_rated_restaurants['rating'].astype('Int64')
df_rated_restaurants.head()
Out[48]:
order_id customer_id restaurant_name cuisine_type cost_of_the_order day_of_the_week rating food_preparation_time delivery_time total_delivery_time
2 1477070 66393 Cafe Habana Mexican 12.23 Weekday 5 23 28 51
3 1477334 106968 Blue Ribbon Fried Chicken American 29.20 Weekend 3 25 15 40
4 1478249 76942 Dirty Bird to Go American 11.59 Weekday 4 25 24 49
5 1477224 147468 Tamarind TriBeCa Indian 25.22 Weekday 3 20 24 44
7 1477859 89574 Barbounia Mediterranean 5.97 Weekday 3 33 30 63
In [49]:
df_rated_restaurants.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 1162 entries, 2 to 1896
Data columns (total 10 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   order_id               1162 non-null   int64  
 1   customer_id            1162 non-null   int64  
 2   restaurant_name        1162 non-null   object 
 3   cuisine_type           1162 non-null   object 
 4   cost_of_the_order      1162 non-null   float64
 5   day_of_the_week        1162 non-null   object 
 6   rating                 1162 non-null   Int64  
 7   food_preparation_time  1162 non-null   int64  
 8   delivery_time          1162 non-null   int64  
 9   total_delivery_time    1162 non-null   int64  
dtypes: Int64(1), float64(1), int64(5), object(3)
memory usage: 101.0+ KB
In [50]:
# Find the number of times each of the rated restaurants(RR) are rated
df_RR_rating_count=df_rated_restaurants.groupby(['restaurant_name'])['rating'].count().reset_index()
df_RR_rating_count = df_RR_rating_count.rename(columns={'rating': 'rating_count'})
df_RR_rating_count
Out[50]:
restaurant_name rating_count
0 'wichcraft 1
1 12 Chairs 2
2 5 Napkin Burger 2
3 67 Burger 1
4 Amma 2
... ... ...
151 Zero Otto Nove 1
152 brgr 1
153 da Umberto 1
154 ilili Restaurant 13
155 indikitch 2

156 rows × 2 columns

In [51]:
# create a new column called 'ave-rating' in the df_RR_rating_count dataframe
df_RR_rating_count['ave_rating']=0
df_RR_rating_count
Out[51]:
restaurant_name rating_count ave_rating
0 'wichcraft 1 0
1 12 Chairs 2 0
2 5 Napkin Burger 2 0
3 67 Burger 1 0
4 Amma 2 0
... ... ... ...
151 Zero Otto Nove 1 0
152 brgr 1 0
153 da Umberto 1 0
154 ilili Restaurant 13 0
155 indikitch 2 0

156 rows × 3 columns

In [52]:
# Find the sum af all the ratings for each rated restaurant
df_RR_rating_sum=df_rated_restaurants.groupby(['restaurant_name'])['rating'].sum().reset_index()
df_RR_rating_sum = df_RR_rating_sum.rename(columns={'rating': 'rating_sum'})
df_RR_rating_sum
Out[52]:
restaurant_name rating_sum
0 'wichcraft 5
1 12 Chairs 9
2 5 Napkin Burger 8
3 67 Burger 5
4 Amma 9
... ... ...
151 Zero Otto Nove 4
152 brgr 3
153 da Umberto 5
154 ilili Restaurant 54
155 indikitch 9

156 rows × 2 columns

In [53]:
# Find the restaurants that meets the criteria for the promotional offer
for i in range(df_RR_rating_count.shape[0]):  # No of rows in the dataframe
    ave_rating=df_RR_rating_sum.rating_sum[i]/df_RR_rating_count.rating_count[i] # Average rating for each rated restaurant
    if df_RR_rating_count.rating_count[i]>50 and ave_rating>4:
        df_RR_rating_count.ave_rating[i]=round(ave_rating,2)
    else:
        df_RR_rating_count.drop([i], axis=0, inplace=True)   # Drop any restaurant that does not meet the criteria from the dataframe.
df_RR_rating_count
Out[53]:
restaurant_name rating_count ave_rating
16 Blue Ribbon Fried Chicken 64 4.33
17 Blue Ribbon Sushi 73 4.22
117 Shake Shack 133 4.28
132 The Meatball Shop 84 4.51

Observations:¶

  • Only 4 out of the 178 restaurants qualify for the promotional offer. These are:

  • Blue Ribbon Fried Chicken with a rating count of 64 and an average rating of 4.33

  • Blue Ribbon Sushi with a rating count of 73 and an average rating of 4.22

  • Shake Shack with a rating count of 133 and an average rating of 4.28

  • The Meatball Shop with a rating count of 84 and an average rating of 4.51

Question 14: The company charges the restaurant 25% on the orders having cost greater than 20 dollars and 15% on the orders having cost greater than 5 dollars. Find the net revenue generated by the company across all orders. [3 marks]¶

In [54]:
# Calculate the revenue generated across all orders
df["revenue"]=0     # Create a new column in the dataframe, initilize it to 0
for i in range(df.shape[0]):   # df.shape[0] is the number of rows in the dataframe
    if df.cost_of_the_order[i]>20:
        df.revenue[i]=df.cost_of_the_order[i]*0.25
    elif df.cost_of_the_order[i]>5:
        df.revenue[i]=df.cost_of_the_order[i]*0.15
    else:
        df.revenue[i]=0
rev_tot=df.revenue.sum()
print('Total revenue is', round(rev_tot, 2), 'dollars')

Total revenue is 6166.3 dollars

Observations:¶

Total revenue generated by the company across all orders is 6166.30 dollars

Question 15: The company wants to analyze the total time required to deliver the food. What percentage of orders take more than 60 minutes to get delivered from the time the order is placed? (The food has to be prepared and then delivered.) [2 marks]¶

In [55]:
# find the number of rows where total delivery time is greater than 60 minutes
count=0
for i in range(df.shape[0]):   # df.shape[0] is the number of rows in the dataframe
    if df.total_delivery_time[i]>60:
        count=count+1
    else:
        count=count 
pct_60=(count/df.shape[0])*100
print(round(pct_60,2),'%')
print(count, 'of the orders placed take more than 60 minutes to deliver')
print('This is about',round(pct_60,2),'percent of the total orders placed')

10.54 %
200 of the orders placed take more than 60 minutes to deliver
This is about 10.54 percent of the total orders placed

Observations:¶

200 out of the total 1898 food order in the dataset took more than 60 minutes to deliver from the time the order was placed to the time the food was delivered. This is about 10.54 percent of the total orders placed and delivered.

Question 16: The company wants to analyze the delivery time of the orders on weekdays and weekends. How does the mean delivery time vary during weekdays and weekends? [2 marks]¶

In [56]:
# Create a new dataframe for food deliveries made on weekdays.
df_weekdays=df[df['day_of_the_week'] == 'Weekday']
ave_wkdy_dt=df_weekdays.delivery_time.mean()
ave_wkdy_tdt=df_weekdays.total_delivery_time.mean()

# Create a new dataframe for food deliveries made on weekends.
df_weekends=df[df['day_of_the_week'] == 'Weekend']
ave_wkend_dt=df_weekends.delivery_time.mean()
ave_wkend_tdt=df_weekends.total_delivery_time.mean()

print(df_weekdays.shape[0], 'deliveries were made on weekdays')
print('Whereas', df_weekends.shape[0], 'deliveries were made on weekends')

print('The average delivery time (after food preparation) on weekdays is', round(ave_wkdy_dt,2),'minutes')
print('Whereas the average delivery time (after food preparation) on weekends is', round(ave_wkend_dt,2),'minutes')

print('The average delivery time (from order placement to delivery) on weekdays is', round(ave_wkdy_tdt,2),'minutes')
print('Whereas the average delivery time (from order placement to delivery) on weekends is', round(ave_wkend_tdt,2),'minutes')

547 deliveries were made on weekdays
Whereas 1351 deliveries were made on weekends
The average delivery time (after food preparation) on weekdays is 28.34 minutes
Whereas the average delivery time (after food preparation) on weekends is 22.47 minutes
The average delivery time (from order placement to delivery) on weekdays is 55.55 minutes
Whereas the average delivery time (from order placement to delivery) on weekends is 49.91 minutes

Observations:¶

Even though there were two and a half times more orders/deliveries made on weekends than weekdays, average food delivery time on weekends is almost 6 minutes faster than what we have on weekdays.

Conclusion and Recommendations¶

Question 17: What are your conclusions from the analysis? What recommendations would you like to share to help improve the business? (You can use cuisine type and feedback ratings to drive your business recommendations.) [6 marks]¶

In [57]:
df_cuisine
Out[57]:
cuisine_type no_of_orders ave_cost_of_each_order total_cost_of_all_orders ave_rating
0 American 584 16.32 9530.78 4.30
1 Japanese 470 16.30 7663.13 4.37
2 Italian 298 16.42 4892.77 4.36
3 Chinese 215 16.31 3505.62 4.34
4 Mexican 77 16.93 1303.85 4.42
5 Indian 73 16.92 1235.14 4.54
6 Middle Eastern 49 18.82 922.21 4.24
7 Mediterranean 46 15.47 711.84 4.22
8 Thai 19 19.21 364.95 4.67
9 French 18 19.79 356.29 4.30
10 Southern 17 19.30 328.11 4.31
11 Korean 13 14.00 182.02 4.11
12 Spanish 12 18.99 227.93 4.83
13 Vietnamese 7 12.88 90.18 4.00
In [58]:
df_rest_count=df.restaurant_name.value_counts().reset_index()
df_rest_count= df_rest_count.rename(columns={'index': 'restaurant_name','restaurant_name':'no_of_orders'})

df_rev_mean=df.groupby(['restaurant_name'])['revenue'].mean().sort_values(ascending = False).reset_index()
df_rev_mean = df_rev_mean.rename(columns={'revenue': 'average_rev_per_order'})

df_rev_sum=df.groupby(['restaurant_name'])['revenue'].sum().sort_values(ascending = False).reset_index()
df_rev_sum = df_rev_sum.rename(columns={'revenue': 'total_revenue_generated'})

df_revenue=pd.merge(df_rest_count, (pd.merge(df_rev_sum, df_rev_mean, on='restaurant_name')), on='restaurant_name')
df_revenue
Out[58]:
restaurant_name no_of_orders total_revenue_generated average_rev_per_order
0 Shake Shack 219 703.61 3.21
1 The Meatball Shop 132 419.83 3.18
2 Blue Ribbon Sushi 119 360.46 3.03
3 Blue Ribbon Fried Chicken 96 340.20 3.54
4 Parm 68 218.56 3.21
... ... ... ... ...
173 Sushi Choshi 1 2.06 2.06
174 Dos Caminos Soho 1 1.90 1.90
175 La Follia 1 1.21 1.21
176 Philippe Chow 1 6.05 6.05
177 'wichcraft 1 1.21 1.21

178 rows × 4 columns

In [59]:
pix = px.bar(df_revenue, x="restaurant_name", y="no_of_orders")
pix.show() 
pix = px.bar(df_revenue, x="restaurant_name", y="total_revenue_generated")
pix.show() 
pix = px.bar(df_revenue, x="restaurant_name", y="average_rev_per_order")
pix.show()
In [60]:
# No of restaurants catering for only 1 order
df_revenue.loc[df_revenue['no_of_orders']==1].shape[0]
Out[60]:
50
In [61]:
# No of restaurants catering for only 2 orders
df_revenue.loc[df_revenue['no_of_orders']==2].shape[0]
Out[61]:
23
In [62]:
# No of restaurants catering for only 1 or 2 orders, whose average revenue generated per order for Foodhub is $2 or less
df20=df_revenue.loc[df_revenue['no_of_orders']<=2]
df21=df20.loc[df20['average_rev_per_order']<=2]
df21
Out[62]:
restaurant_name no_of_orders total_revenue_generated average_rev_per_order
108 Zero Otto Nove 2 2.71 1.35
110 brgr 2 2.97 1.48
113 Olive Garden 2 3.43 1.71
117 Bukhara Grill 2 3.56 1.78
119 Haandi 2 2.82 1.41
125 Xe May Sandwich Shop 2 3.30 1.65
127 Mira Sushi 2 3.73 1.86
131 Frank Restaurant 1 1.81 1.81
132 Hiroko's Place 1 1.96 1.96
133 Rye House 1 1.82 1.82
134 Lamarca Pasta 1 1.38 1.38
138 Hampton Chutney Co. 1 1.86 1.86
139 Wa Jeal 1 1.01 1.01
140 Hunan Manor 1 0.86 0.86
144 Market Table 1 1.02 1.02
147 Coppola's East 1 1.96 1.96
148 Kori Restaurant and Bar 1 1.83 1.83
149 Woorijip 1 1.24 1.24
151 The MasalaWala 1 1.82 1.82
153 Big Wong Restaurant Œ_¤¾Ñ¼ 1 0.89 0.89
159 Samurai Mama 1 1.75 1.75
161 Balade 1 1.38 1.38
166 Gaia Italian Cafe 1 1.32 1.32
167 Chola Eclectic Indian Cuisine 1 1.31 1.31
171 Alidoro 1 1.89 1.89
172 Cipriani Le Specialita 1 0.89 0.89
174 Dos Caminos Soho 1 1.90 1.90
175 La Follia 1 1.21 1.21
177 'wichcraft 1 1.21 1.21
In [63]:
# Total revenue generated for Foodhub for servicing the restaurants above
df21.total_revenue_generated.sum()
Out[63]:
54.807
In [64]:
# Number of restaurants with only 1 or 2 orders whose average generated revenue is $2 or less
df21.shape[0]
Out[64]:
29

Conclusions:¶

  • We have analyzed the dataset given by the foodhub company. It contains a total of 1898 rows and 9 columns. the dataset does not contain any missing value except for the fact that almost 40 percent of the customers fail to rate the service being provided on receipt of their order.
  • It is noteworthy to point out that no customer ever gave Foodhub services any rating lower than 3, which is the average rating on a scale of 1 to 5. In fact about 50% of the customers gave theservice being provided a rating of 5, while another 33 percent gave the services being provided a rating of 4.
  • There are 178 restaurants in FoodHub's database. FoodHub provides its customers a choice of 14 different cuisine types on their database. The "go to" restaurant is Shake shack which caters for a disproportionate quantity (more than 11 percent) of the total orders placed and delivered.
  • It is also observed that American cuisines are the favoured dish of most of the customers with more than 30 percent of the orders being placed and delivered being for American type food.
  • More than 71 percent of the business done by FoodHub are done on weekends with weekday orders responsible for less than 29 percent of the services delivered.
  • Despite the fact that a disproportionate volume of the orders catered for are done on weekends, food delivery is actually faster on weekends than on weekdays. from the analysis, it was discovered that food preparation takes approximately the same amount of time both on weekdays and weekends, but actual delivery after the food has been prepared is slower by about 6 minutes on weekdays. This is likely due to the fact that there is often more traffic on weekdays than on weekends.

Recommendations:¶

  • Delivery times are slower on weekdays than weekends. This is actually the time people may be pressed for time and the delay in delivery may actually be responsible for the lower volume of orders received on weekdays. Since there is little or nothing that we can do to overcome the constraint of heavier weekday traffic, we may want to see what could be done to speed up food preparation times on weekdays.
  • A close perusal of the revenue dataset (df_revenue) shows that patronage of some of the restaurants are very low . 50 out of the 178 restaurants (about 28 percent) catered for just 1 order according to the dataset. Another 23 out of the 178 (almost 13 percent) catered for just two orders. There might be a need to prune the restaurant list on the App depending on the logistics of servicing each order. For example, if we spend 2 dollars to service each order, then it may be pointless having the 29 restaurants listed above on the App as the total revenue generated through them is not enough to support the logistics of fulfilling the orders emanating from them.